Symptom
'netstack' process crash on a Nexus switch due to mbuf memory leak while processing MPLS packets
Conditions
'netstack' process reloads, generally with a core file, and other services may be affected. Additionally, the following errors may be seen following the event -
%NETSTACK-2-MPULLUP: netstack [PID #] p_ip_output: m_pullup failed for IP, error Resource temporarily unavailable
%NETSTACK-2-MPULLUP: netstack [PID #] p_ip_output: m_pullup failed for IP, error No route to host
%NETSTACK-2-MPREPEND: netstack [PID #] udp_output: m_prepend failed for connection
%NETSTACK-2-MPULLUP: netstack [PID #] p_ip_output: m_pullup failed for IP, error No route to host
%NETSTACK-2-MPULLUP: netstack [PID #] p_ip_output: m_pullup failed for IP, error Operation not permitted
This will be seen with MPLS configured and operational on the box.
Workaround
Kill netstack process from kernel or reload the vdc/switch
Further Problem Description
memory leak can lead to no buffer errors as well:
NETSTACK-3-NO_MBUF netstack [] Failed to allocate mbuf (cluster) to read from Inband
